In this exercise, you'll migrate an example notebook (fitting a PyTorch CNN model on the MNIST Digits sample dataset) into the SageMaker data science workflow.
To get started, clone this repository into a SageMaker Notebook instance (any instance type will do) and fire up the Instructions.ipynb notebook!
This practice exercise is intended to be delivered with in-person support, and assumes you:
- Have had a high-level introduction to the SageMaker workflow, and:
- Are familiar with using the AWS Console to access Amazon SageMaker and Amazon S3
- Are familiar with configuring SageMaker Notebook Instance Execution Roles with appropriate Amazon S3 access
If that doesn't sound like you, you might prefer to check out:
- The official Introductory Amazon SageMaker Tutorial
- The "Get Started with the Amazon SageMaker Console" page in the Amazon SageMaker Developer Guide